When elements are simultaneously added to lakes, experimentally or by accident, their ratios in the water phase and in bottom sediments can change with time due to differential partitioning between solution and suspended particles or sediments. A number of equations are developed to show the change of ratio with time in water and sediments assuming simultaneous pulse inputs followed by a range of combinations of loss processes from solution (i.e. hydraulic losses, sorption to particles followed by settling, and diffusion into the sediments). The pattern of events is discussed both for pulse events with specific limiting assumptions and for combined continuous and pulse inputs. The models show that elemental ratios in sediments are generally less sensitive indicators of differential partitioning than are elemental ratios in water. For lakes with long residence times, the long-term elemental ratio in the sediments does not differ from that in the initial spike to the water column, but for short residence times, it is directly dependent on the ratio of either partition or diffusion coefficients. The models are used to interpret Ru : Cs ratios measured in the water and sediments of Esthwaite Water subsequent to the pulse input of Chernobyl fallout. The ratios can be explained by assuming nuclides were lost either by flushing and sorption or by flushing, sorption, and diffusion. The process combination of flushing and diffusion is incompatible with the observed constant ratios.